earthing conductor sizes

[JohnW2]

I think a fault inside the CU but before any protective device is probablly more likely than a fault between meter and CU. Particularly with a metal cased CU.

Agreed - to be clearer, I really should have written "... an L-E fault between DNO fuse and the OPDs within the CU" - but I would imagine that, even with a metal CU, that would be very rare.

However, none of these side issue raised by EFLI have any relevance to the simple point I was trying to make in response to him having ;posted a copy of 543.1.3 - that one cannot undertake the adiabatic calculation described in that reg without referring to manufacturer's I²t/I data for the relevant OPD (be that the DNO's fuse, an MCB or whatever), since the generally available t/I curves (e.g. those in BS7671) simply don't allow one to ascertain the disconnection time (hence I²t) at likely PFCs.

Kind Regards, John

 

[EFLImpudence]

I've never really understood what "Ia" (alone) represents, since it only seems to have meaning when viewed in conjunction with some specified disconnection time - since it is defined as "the current causing operation of the protective device within the specified time".

Whatever it represents, is not its numerical value what you called NxIn?
Just as in the supplementary bonding requirements that you have discredited, i.e. R≤50/Ia.

Yes, of course, but I think you're overthinking the statement I made. I was merely pointing out that t/I curves are essentially vertical at the ('worst case') magnetic trip threshold for an MCB - and, as you say, a fault current could be much higher than that. That's why, as I said, one cannot determine the disconnection time (hence I²t, for an adiabatic calculation) at the PFC from the standard t/I curves. Much the same is true of fuses, even though the curves never become 'vertical' - indeed, the curves for fuses (as well as MCBs) in BS7671 all stop at 100 ms, so one cannot determine a disconnection time less than that.

If the CPC csa is deemed satisfactory using 0.1s then it will also be so at shorter times.
Don't forget that the Earthing Conductor has a minimum of 6mm² (or larger depending on the earthing method).

Are you suggesting that, contrary to what I said, you have some way of determining I²t for an diabatic calculation without access to I²t/I curves for the relevant OPD (which curves, as I said, do not exist in BS7671 or any associated documents)?

No, but if it is necessary to determine the actual time then it will have to be determined.

That's true if you feel the need to consider the (very improbable) possibility of an L-E fault between DNO fuse and CU. If you were prepared to ignore that remote possibility, then the relevant OPD would presumably be the highest-rated of the MCBs in the CUI?

Nevertheless that is how the csa of the Earthing Conductor is determined (apart from specific instruction).

 

[JohnW2]

Whatever it represents, is not its numerical value what you called NxIn?

It would IF "the specified time" (per definition of Ia) corresponded to the (worst case) magnetic disconnection time of the MCB. It is because of that 'non-precise' definition of Ia that I personally never use it.

If the CPC csa is deemed satisfactory using 0.1s then it will also be so at shorter times.

That is obviously true but, at least in theory, 0.1s could be an excessively 'demanding' figure to use (particularly for an MCB, for which the true value of t would presumably be a lot less than 0.1s). However, having said that, as you go on to say ...

Don't forget that the Earthing Conductor has a minimum of 6mm² (or larger depending on the earthing method).

In most practical (domestic) situations, that really precludes the need for any calculation. To require a CSA >6mm² would, even for a 0.1s disconnection time, require a PFC of about 2,182 A (i.e. a Ze of about 0.1 Ω), which is much greater than any I've personally ever seen (in my limited experience) in a domestic property. In the case of a TN-C-S supply, with a minimum earthing conductor CSA or 10mm, the PFC would have too be at least about 3,637 A (Ze about 0.06 Ω), even further beyond my experience.

No, but if it is necessary to determine the actual time then it will have to be determined.

That was my point. However, as above, and as you have implied, an accurate ('correct') adiabatic calculation will not usually be needed, since a 'lower bound' of required CSA (assuming a disconnection time of 0.1s) is adequate.

Having said that, one probably does have to go through the motions of undertaking the adiabatic calculation and get that 'lower bound' - since, if one doesn't one is probably stuck with Table 54.7 which will usually result in a minimum CSA of 16mm² for main bonding and earthing conductors. I say 'probably', because it's all a bit confusing, and one might be able to use 10mm² (the minimum main bonding CSA) ion teh case of TN-C-S.

Nevertheless that is how the csa of the Earthing Conductor is determined (apart from specific instruction).

Fair enough.

KInd Regards, John